Aim. To assess and compare the indexes of clinical outcomes of stroke comprehensively and to explore the way to improve them so as to evaluate the efficacy of traditional Chinese medicine (TCM) therapies on stroke more comprehensively, dynamically and scientifically. Methods. 245 patients with stroke, 159 males and 86 females, aged (66 ± 10) years, 79 with a course less than 1 month, 47 with a course of 1-3 months, and 57 with a course of 3-6 months, 17% being illiterate, 38% having received the education of primary school level, 32% having the education of middle school level, and 13% having received higher education, were assessed by cross-sectional study combined with prospective follow-up to measure the neurological defect scale(NDS), living ability, activities of daily living (ADL, Barthel index). TCM diagnostic and evaluation criteria of stroke, scales of TCM syndrome related indexes. Quality of life was measured with Medical Outcomes Study 36-Item Short-Form Health Survey(MOS SF-36 Health Survey) and quality of life index(QOI). Some patients were re-evaluated 2 months after the first evaluation. The methods of multivariate relative regression and linear regression, factor analysis and principle component analysis, Cronbach's αcoefficients, and the estimation of responsiveness were used to evaluate the validity, reliability and responsiveness of the indexes. Results. The correlation of NDS score with score based on TCM criteria was higher than 0.9. The reliability, validity, and effect size of the modified neurological functioning scale were 8.5% higher than those of the original scales. The Barthel index showed a "floor effect" in evaluating mild stroke patients and several items of the physical functioning scale of SF-36 did not fit for evaluating stroke patients. These two scales could be combined to form a new scale. Many famous foreign mental health scales, social functioning and physical role scales, such as Mental Health Inventory (MHI-5) showed shortages in this study. Conclusion. The present indexes for assessing the clinical outcomes of stroke still need improvements. Clinical outcomes of stroke should be assessed comprehensively and dynamically with various indexes, according to different stage and severity on the basis of systematic and strict assessment and improvements of the present indexes.